JP2002215236A - Controller for travel of unmanned vehicle - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an unmanned vehicle traveling controller capable of maintaining the accuracy of traveling guide at a high level even when the position measuring accuracy of GPS navigation or autonomous navigation is deteriorated or position measurement is disabled. SOLUTION: The unmanned vehicle traveling controller provided with a GPS receiver (13) for receiving a GPS signal and measuring the position of a vehicle, an autonomous navigation computing element (14) for measuring the position and azimuth of the vehicle on the basis of the traveling direction and distance of the vehicle, a position measuring part (11) for calculating the current position and azimuth of the vehicle on the basis of the position measuring results of the receiver (13) and the computing element (14), and a traveling control part (17) for controlling the traveling of the vehicle on the basis of a compared result between a previously set traveling course (4) and the calculated current position and azimuth is also provided with a road side band distance measuring instrument (15) for measuring a distance from the vehicle up to a road side band (5) formed on the side of the traveling course (4). The position measuring part (11) finds out the current position and the azimuth by correcting at least either one of the measuring position of the GPS receiver (13) and/or the measuring position and azimuth of the computing element (14) on the basis of the road side band distance measured by the measuring instrument (15).

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a traveling control device for an unmanned vehicle that automatically travels on a predetermined traveling route in a mine, a quarry, or the like.

[0002]

2. Description of the Related Art Conventionally, as a traveling system of an unmanned vehicle that automatically travels on a predetermined traveling route, a satellite positioning navigation (hereinafter, referred to as an “absolute position”) that uses a GPS receiver to determine the absolute position of the unmanned vehicle.
GPS navigation) and autonomous navigation that detects the traveling direction and traveling distance of the unmanned vehicle and estimates and calculates the relative position and azimuth based on the absolute position using an internal sensor. It is well known that the vehicle travels with improved position measurement accuracy by complementation. The autonomous navigation includes dead reckoning (a method of estimating the traveling distance using a speed sensor or a moving distance sensor and estimating the direction using a gyro or a geomagnetic sensor to estimate the position and direction of the vehicle) and And inertial navigation (navigation for estimating the position and orientation of a vehicle using a gyro and an acceleration sensor) are well known. Hereinafter, these are collectively called autonomous navigation.

FIG. 12 is a diagram illustrating a traveling system of an unmanned vehicle. In the figure, a fixed base station 1 controls the control of an unmanned vehicle 10 and a control communication antenna 2
Through b, travel route data and a vehicle control command are transmitted to the unmanned vehicle 10, and a vehicle state signal is received.

[0004] The unmanned vehicle 10 has a position measuring unit and a traveling control unit. The position measurement unit obtains the absolute coordinate value of the current position of the vehicle itself based on the position information from the plurality of satellites 3a, 3e received by the GPS receiver via the GPS antenna 12a (GPS navigation). At this time, the position measurement unit obtains a relative position and an azimuth from the absolute coordinate position based on the GPS based on a traveling direction and a traveling distance, and sequentially estimates and calculates a current position and an azimuth (autonomous navigation). Then, after traveling for a predetermined GPS positioning sampling time, the current position and the direction based on the relative position and the direction are corrected based on the absolute position newly measured by the GPS. The traveling control unit calculates the traveling direction based on a comparison result between the preset position and orientation (traveling direction) of the traveling route and the current position and orientation, and travels at a predetermined speed in the traveling direction. The steering and the vehicle speed are controlled so that The above process is repeated to guide the unmanned vehicle along the traveling route.

In the GPS navigation described above, the calculation time from the start of positioning, the communication delay from the GPS receiver to the final position measurement unit, and the timing time interval for outputting the positioning data are determined by the position in the autonomous navigation. There is a disadvantage that it is slower than the calculation time. Further, the reception environment of the satellites 3a to 3e, that is, the visibility of the sky, whether the sky above the unmanned vehicle is in a state where GPS reception can be normally performed due to surrounding rocks, forests, structures, and the like, the satellite arrangement, and Due to reception environment factors such as multipath when radio waves reflected from surrounding reflectors are received, DGPS (so-called differential system G
In the case of PS), when the correction information communication from the fixed base station 1 is delayed due to various factors such as a mountain, a quay, a forest, a structure, and a radio wave environment, the positioning accuracy is deteriorated or the positioning becomes impossible. There is also a known problem.

[0006] On the other hand, autonomous navigation is used to compensate for the drawback. For example, a relative distance is calculated by calculating a traveling distance in a short time at predetermined time intervals based on measurement values of a speed sensor and a direction sensor. So you can measure in real time. However, since sensor errors such as a speed sensor and an azimuth sensor are accumulated and there is a problem of deterioration of positioning accuracy, the G
Correction is performed at predetermined time intervals by PS navigation to maintain positional accuracy.

[0007]

However, the conventional traveling system for an unmanned vehicle has the following problems. (1) In the GPS navigation, as described above, the positioning accuracy may be deteriorated or the positioning may not be performed due to the deterioration of the satellite reception environment or the deterioration of the reception environment of the correction data from the base station in the case of DGPS. . In this case, the accuracy of the current position and the current direction obtained based on the absolute position by the GPS navigation and the relative position and the direction by the autonomous navigation deteriorates. The accuracy of the current bearing deteriorates. For this reason, guidance accuracy of the unmanned vehicle is reduced, which hinders vehicle control. (2) When the speed sensor or the direction sensor related to the autonomous navigation fails, if only the GPS navigation is used, the time interval of the positioning data output by the GPS navigation is slow as described above. The traveling guidance accuracy during this time is reduced. In addition, if only GPS navigation is used, speed accuracy and azimuth accuracy deteriorate at low speeds, and a GPS antenna trajectory (corresponding to GPS azimuth) during turning.
Is different from the attitude direction of the vehicle, the accuracy of positioning (position and direction) is degraded, and the traveling guidance accuracy is reduced. At this time, if the positioning (position and orientation) accuracy is reduced due to the deterioration of the reception environment or communication environment of the GPS navigation, the traveling guidance accuracy is further reduced. For this reason,
In the case of a failure that disables autonomous navigation, it hinders vehicle control and reduces the operating rate of unmanned vehicles. However, there is a strong demand for a highly reliable traveling system that can reliably perform guided traveling even in the above case.

The present invention has been made in view of the above-mentioned problems, and is capable of maintaining high driving guidance accuracy even when the positioning accuracy of the GPS navigation or the autonomous navigation is deteriorated or the positioning cannot be performed. It is an object of the present invention to provide a traveling control device for a vehicle.

[0009]

In order to achieve the above object, a first invention is to provide a GPS receiver which receives a GPS signal to measure an absolute coordinate position of a vehicle, and a traveling direction of the vehicle. And an autonomous navigation calculator that measures the relative coordinate position and azimuth from a specific position of the vehicle based on the measurement result, and a GPS receiver and an autonomous navigation calculator based on the respective positioning results of the vehicle. A position measuring unit that calculates the current position and direction, and a vehicle that compares the position and direction of the traveling route set in advance with the current position and direction calculated by the position measuring unit, and reduces the respective deviation values. A travel control unit for controlling the travel of the unmanned vehicle, comprising: a roadside zone distance measuring device that measures a distance from the vehicle to a roadside zone provided on a side of the travel route; The measuring unit corrects at least one of the positioning position of the GPS receiver and / or the positioning position and the azimuth of the autonomous navigation calculator based on the roadside zone distance measured by the roadside zone distance measuring device, and corrects the current position and the current position. It is configured to obtain the direction.

According to the first invention, the roadside belt (road shoulder, signboard,
(Composed of a reflector, etc.) based on the measured value of the distance to the current position and the azimuth by correcting at least one of the positioning position by the GPS navigation and / or the positioning position and the azimuth by the autonomous navigation. Since the calculation is performed, a decrease in the measurement accuracy of the position and the direction due to the GPS positioning error and the positioning error of the autonomous navigation is reduced, and the vehicle position and the direction with higher reliability can be measured. In addition, even if positioning by GPS navigation or autonomous navigation becomes impossible or abnormalities in positioning data are recognized, positioning (position and direction) data by the normal GPS navigation or autonomous navigation and measured values of roadside zone distance Thus, a reliable current position and direction can be obtained based on the above, so that guided driving can be continued and the operation rate of the unmanned vehicle can be increased. As a result, a highly reliable traveling system for unmanned vehicles can be configured.

According to a second aspect based on the first aspect, the position measuring section determines the stability of the roadside zone from the measured value of the roadside zone distance, and performs correction based on the roadside zone distance when the stability is good. I'm trying to do it.

According to the second aspect of the present invention, the roadside zone is unstable, that is, the roadside zone is discontinuously interrupted in an area which is not set in advance or an area where the change rate is larger than a predetermined change rate. It is determined that some abnormality has occurred, and the roadside zone distance measured in an area having such an abnormality is regarded as abnormal data. Here, as the topographical information of the roadside zone, the roadside zone distance change rate with respect to the installation range of the road shoulder and the traveling route, the installation coordinates and the shape of the signboard and the reflector, and the like are set in advance. Therefore, when the roadside zone is stable,
The correction is performed based on the roadside belt distance. Thereby, when the roadside zone is unstable, it is possible to prevent the positioning (position and azimuth) accuracy from lowering due to the correction, continue the automatic driving, and increase the operating rate of the unmanned vehicle.

A third invention is based on the first or second invention,
The correction based on the roadside zone distance of the position measuring unit is performed by a sensor corresponding to the positioning accuracy estimated value of the GPS receiver and / or the positioning accuracy estimated value of the autonomous navigation calculator and the positioning accuracy estimated value of the roadside zone distance measuring device. It is done by fusion.

According to the third aspect of the present invention, the corrected position and the corrected azimuth are calculated by the sensor fusion corresponding to the estimated values of the positioning accuracy of the GPS receiver, the autonomous navigation calculator, and the roadside distance measuring device, so that the reliability is improved. It is possible to obtain a position with high probability. In addition, the positioning accuracy estimation value is a GPS receiver, an autonomous navigation calculator, a roadside zone distance measuring device, and each of the recognized sensor information and recognized sensor information, and is estimated from empirical judgment. It may be estimated in real time, or may be set to a predetermined value in advance.

[0015]

Embodiments of the present invention will be described below in detail with reference to the drawings.

The basic configuration of the traveling system for an unmanned vehicle according to the present invention is the same as that shown in FIG. Here, the position (absolute coordinates) of the fixed base station 1 is accurately measured in advance. The fixed base station 1 calculates absolute coordinate values by GPS based on position information from a plurality of satellites 3a and 3e received by a GPS receiver via a GPS antenna 2a, and calculates the absolute coordinate values and the calculated absolute coordinate values in advance. A positioning error parameter by GPS at each calculation time point (a parameter representing a common error of each receiver, for example, receiving distance information from a plurality of satellites 3a to 3e) Time delay data of the internal clock of the receiver, time information data, satellite orbit error data, etc.). Then, the obtained GPS positioning error parameter is transmitted as correction data to the unmanned vehicle 10 via the control communication antenna 2b.

Referring to FIG. 1, the configuration of a control device for an unmanned vehicle according to the present invention will be described. The GPS receiver 13 receives position information from a plurality of GPS satellites via the GPS antenna 12a, and receives the absolute position, direction, and speed of the vehicle (in this case, G
(A GPS attitude angle indicating the moving direction of the PS antenna 12a and a GPS speed indicating the moving speed). At this time, the absolute coordinate value, azimuth, and speed measured by the vehicle itself are corrected based on the correction data received from the fixed base station 1 via the control communication antenna 12b (so-called differential GPS). The GPS receiver 13 has its own failure diagnosis function.
Analysis of a communication abnormality signal included in the GPS communication information, that is, a communication abnormality signal detected by a GPS satellite, detection of reception data abnormality on the GPS receiver 13 side, and the like are performed.

The autonomous navigation computing unit 14 measures a relative movement position from a position serving as a base point, and calculates a relative position and an azimuth (in this case, a traveling direction) by autonomous navigation. In the present embodiment, an example of dead reckoning is shown, which has a gyro 21 for measuring the traveling direction of the unmanned vehicle and a speed sensor 22 for measuring the traveling speed, and is calculated from the measured traveling direction and the traveling speed. The relative position and azimuth are obtained from the traveling distance. In the case of inertial navigation, as with the general configuration, a gyro platform that always maintains an equilibrium state in the direction of gravity is provided,
This platform is equipped with an acceleration sensor, which integrates the acceleration signal to obtain a speed signal, further calculates the speed signal to obtain the travel distance, detects the traveling direction with a gyro, and uses these data as in dead reckoning navigation. Are determined. Although the following description describes dead reckoning navigation, inertial navigation can be similarly configured. The speed sensor 22 is constituted by, for example, a rotation sensor for detecting the number of rotations of the wheels, and is provided at at least one of the four wheels (for example, at the three positions of the left and right front wheels and the rear wheels). Further, the autonomous navigation computing unit 14 has its own failure diagnosis function, and analyzes an abnormal input data from the gyro 21 and an abnormal signal, an abnormal input data from the speed sensor 22 (for example, an abnormal change in the detection signal, Failure diagnosis is performed by comparing the signals from the speed sensors 22 when the speed sensors 22 are provided on the left and right front wheels and one rear wheel, and analyzing a disconnection detection signal.

The roadside zone distance measuring device 15 measures the distance from the unmanned vehicle 10 to the roadside zone of the traveling route (such as a bank or a wall or a reflector representing the roadside zone) in a non-contact manner. It has a total 23, an ultrasonic distance meter, and the like. The roadside distance measuring device 15 has its own failure diagnosis function, and analyzes, for example, a change abnormality in a measurement signal.

The position measuring unit 11 is a positioning data (position and direction) of each positioning module (ie, GPS receiver 13, autonomous navigation calculator 14, and roadside distance measuring device 15). ), Weighting according to the reliability of each positioning data is taken into account and averaged to calculate a more reliable and accurate position and orientation of the unmanned vehicle. Hereinafter, positioning based on averaging in consideration of such weights is referred to as sensor fusion.
That is, based on a failure diagnosis result signal from each of the positioning modules 13, 14, and 15, or a deviation (data inconsistency or data abnormality) between the positioning data of each of the positioning modules 13, 14, and 15, there is no abnormality. At that time, positioning is performed by sensor fusion according to the reliability of each positioning data, that is, the estimated value of positioning accuracy (hereinafter, simply referred to as positioning accuracy), and when there is an abnormality in any of the positioning modules, Excluding the positioning data, the other positioning data are complemented with each other by sensor fusion according to the positioning accuracy (reliability), thereby obtaining a more accurate position and orientation of the unmanned vehicle. Note that the positioning accuracy of each of the positioning modules 13, 14, and 15 is estimated from the result of positioning, the information of each recognized sensor, and empirical judgment, and the estimated value is obtained in real time during actual processing. Estimation calculation may be performed, or a predetermined value may be set in advance.

Incidentally, as a method for detecting the above data abnormality,
For example, terrain information along the traveling route is input from the traffic control unit 16, and the input terrain information and the roadside distance measuring device 1 are input.
Data abnormality determination is performed based on the difference value information from the positioning data from No. 5. Then, in order to allow the unmanned vehicle to travel based on the obtained accurate positioning data, the positioning data is transmitted to the traffic control unit 16, the traveling control unit 17, and the safety management unit 1, respectively.
9 is output.

The traffic control unit 16 stores travel route data C
D and the terrain information GI are stored. During automatic driving, the stored traveling route data CD is stored in the travel control unit 17 and the position measurement unit 11 and the terrain information GI is stored in the position measurement unit 1.
1 is output. The terrain information GI is information relating to a roadside zone installed along the traveling route, and details will be described later. The travel route data CD may be created by teaching the vehicle while actually traveling manually, or may be created offline in advance based on survey data of the travel route from the fixed base station 1 in the control communication. You may make it receive via the antenna 12b. Further, the traveling route data CD created by the above teaching is edited according to the distance to the roadside belt, that is, in association with the roadside belt distance,
It may be modified. Teaching is GP
This is performed when the condition of S is favorable, and the data of the position, azimuth (running direction) and vehicle speed by GPS when the vehicle actually travels along the running route are stored at predetermined time intervals. At this time, as the position and direction data, the position and the direction by the GPS and the position and the direction by the autonomous navigation are determined according to the reliability of the respective positioning data, that is, the interruption time of the GPS communication signal, the GPS positioning error, and the autonomous navigation. It is preferable to use the position and orientation estimated by sensor fusion according to the magnitude of the measurement error and the like.

The traveling control unit 17 controls various actuators 18 for controlling steering and speed so that the unmanned vehicle travels according to the traveling route data. The actuator 18 includes, for example, a steering control valve for controlling a steering cylinder for steering, a brake control valve for controlling a brake and a retarder, and a motor for controlling a swing angle of an engine speed control lever. Travel control unit 1
7 is a target position and a target azimuth on the traveling route and a position measuring unit 1
The driving direction and the traveling distance are calculated so that the respective deviations from the current position and the azimuth input from 1 are reduced, and each control command is output to the actuator 18 based on the calculated traveling direction and the traveling distance, and the unmanned vehicle To control the automatic driving of the vehicle.

The safety management section 19 has a function of detecting an obstacle (penetration of a person or a car, falling rocks, etc.) on a traveling course ahead of the unmanned vehicle based on the current position data from the position measurement section 11. When an object is detected, the unmanned vehicle is safely stopped, or the vehicle is driven around an obstacle. When the position measurement system or the control system has some unexpected abnormality (that is, it cannot be detected by itself), a roadside zone or an oncoming vehicle which should not be on the course is detected as an obstacle. Stop the vehicle immediately.

Next, the above terrain information will be described in detail. The current position of the unmanned vehicle is estimated based on the distance between the roadside zone and the traveling unmanned vehicle measured by the roadside zone distance measuring device 15, and the estimated position data and another GPS receiver 13 or autonomous navigation calculator 14 In order to obtain an accurate current position and azimuth by performing sensor fusion with the positioning data according to the above, accurate distance information from the traveling route corresponding to each traveling position of the traveling route to the roadside zone is required. The terrain information GI is roadside zone information for obtaining this accurate distance information. The terrain information GI is determined by conditions such as the width of the traveling route, terrain information indicating the presence or absence and continuity of a bank beside the route, and the traveling direction distance of a traveling area in which visibility over the sky is poor and sensor fusion is required at the roadside zone distance. The following various data setting methods can be considered, each of which has different data contents.

Setting method of the terrain information GI (1) When the traveling road width is uniform and the bank beside the traveling road is well maintained, as shown in FIG. 2, traveling route 4 (substantially center line of the road width) Since the distance between the vehicle and the roadside zone 5 is constant, a reference value L0 of the distance L between the unmanned vehicle 10 and the roadside zone 5 is set in advance. The current position is estimated based on the difference between the reference distance L0 and the value measured by the roadside distance measuring device 15.

(2) When the running path width is uneven and the bank is well-maintained, but the distance between the running path and the bank is uneven (when the reference distance L0 is not constant), as shown in FIG. Since the distance between the traveling route 4 and the roadside zone 5 changes according to the traveling route 4, it is necessary to sequentially calculate the reference distance L0 according to the traveling route 4. In this case, the inclination of the bank with respect to the travel route 4 (the rate of change of the roadside belt distance) is also stored as the topographic information GI. Terrain information G
I has several setting methods, for example, as follows. 2-1) A method of storing the coordinate data of the precisely measured bank on the unmanned vehicle 10 side according to the travel route 4 The reference distance L0 is sequentially obtained and obtained from the distance between the stored bank coordinates and the travel route 4. The current position is estimated from the difference between the reference distance L0 and the actually measured distance data L. Further, the inclination of the bank with respect to the travel route 4 (the rate of change of the roadside zone distance) is calculated based on the distance between the bank coordinates and the travel route 4. In addition, it is also possible to compare the bank coordinates, the own vehicle position, and the traveling route 4 to detect whether there is an abnormality such as overlapping, approaching, and deviation (contradiction) of coordinate data. 2-2) The vehicle control system at the base station calculates the distance between the precisely measured bank coordinate data and the travel route 4, and the unmanned vehicle 10 receives from the vehicle control system according to the travel route 4. This method has an advantage that the calculation processing time on the unmanned vehicle 10 side is shortened. 2-3) The distance L to the bank is measured according to the traveling route 4 at the time of the teaching of the traveling route 4, and the teaching data of the traveling route 4 and the measured roadside belt distance L are both determined on the vehicle control system side. Memorize, during automatic travel, travel route 4
To receive the roadside distance L according to

(3) In the case where an area without a bank has been maintained, predetermined reflectors such as reflection poles and signboards are installed at regular intervals on the side of the traveling road, and as the topographical information GI, the coordinates of the reflector installation positions or the reflectors are set. The distance between the reflection object and the travel route 4 is stored. At the time of automatic traveling, the reference distance L0 between the unmanned vehicle 10 and the roadside zone 5 (reflecting object) is calculated from the stored coordinates of the reflecting object and the coordinates of the traveling route 4 or the distance between the stored reflecting object and the traveling route 4. , The reference distance L0 is calculated, and the current position of the unmanned vehicle 10 is estimated based on the reference distance L0.
The reflector installation coordinates may be obtained by surveying, or the reflector position may be detected and stored at the same time as the teaching operation. Further, the distance between the reflection object and the travel route 4 may be similarly calculated from the survey result, or may be detected and stored simultaneously with the teaching.

(4) A case where position correction in the vehicle running direction is required even if there is a repaired bank. In this case, a reflection object is installed at a fixed interval, or a point where the running direction position correction is required (a turning point). , Etc.). The coordinates of these reflectors are accurately measured in advance,
It is stored as terrain information GI.

Next, the sensor fusion function of the position measuring section 11 will be described in detail. (Sensor fusion between GPS and autonomous navigation) GPS
Both the receiver 13 and the autonomous navigation calculator 14 have no failure,
If there is no abnormality in each positioning data, the current position and the azimuth are calculated by the sensor fusion of the positioning data of both. That is, as shown in FIG. 4, while traveling, the GPS is referred to by reference to the GPS absolute coordinates and azimuth measured by the GPS receiver 13 at predetermined time intervals and the correction data from the fixed base station 1 at this time. Positioning data (position and direction) is obtained by using the GPS, and the measurement accuracy of the positioning data by the GPS is compared with the measurement accuracy of the positioning data by the autonomous navigation so far, and the weight according to the accuracy (reliability) of each positioning data is obtained. (E.g., 0.8 and 0.2) are multiplied to obtain an average value, which is set as a corrected current position, and the azimuth is corrected based on the correction amount of the current position. When the measurement accuracy of the azimuth data by the autonomous navigation is high, the sensor fusion may be performed using only the position data among the positioning data by the GPS. The traveling control unit calculates a deviation (distance) between the corrected current position and a target position on the traveling route 4.
α, the steering amount is determined in a direction to reduce the deviation between the actual posture angle of the vehicle (that is, the traveling direction) and the target posture angle of the traveling route 4, and the unmanned vehicle 10 travels. The present position and direction are estimated and calculated from the relative position based on the corrected current position and direction by the dead reckoning navigation. The above-described sensor fusion between GPS navigation and autonomous navigation is repeated.

(Detection of abnormality in sensor fusion between GPS and autonomous navigation) The GPS receiver 13 and the autonomous navigation arithmetic unit 14 are normal, but if an abnormality (deviation) is found between the positioning data, the GPS reception is performed. The current position and azimuth are estimated from the sensor fusion of the respective positioning data of the device 13, the autonomous navigation calculator 14, and the roadside zone distance measuring device 15. That is, the current position obtained by the sensor fusion of each positioning data by GPS and autonomous navigation is further corrected by the sensor fusion with the position data based on the roadside distance to obtain a more reliable current position, and the current position is corrected. The azimuth by the GPS navigation and the autonomous navigation is corrected based on the amount, and a more reliable azimuth is obtained. Here, as the above-described data abnormality detection method, for example, as shown in FIG. 5, the data is set or obtained based on the measured value of the roadside distance L measured by the normal roadside distance measuring device 15 and the terrain information GI. When the difference value from the reference distance L0 exceeds a predetermined value (that is, when the difference value becomes extremely small or excessively large), it can be determined that there is an abnormality. Alternatively, when the difference value between the respective pieces of positioning data obtained by different positioning methods is equal to or more than a predetermined value, there may be a large difference between the data and the data may be regarded as abnormal. However, the stability of the data of the reference distance L0 and the actual measurement distance of the roadside zone is determined, and the continuity is surely recognized and the bank shape is detected stably. An abnormality detection determination is made. Naturally, when a failure of the roadside belt distance measuring device 15 occurs, the abnormality diagnosis based on the magnitude of the difference between the above-described actual measurement distance of the roadside belt and the reference distance L0, and the sensor fusion based on the roadside belt distance, Shall not be performed.

(Sensor fusion of roadside zone measurement and autonomous navigation when GPS is abnormal) If abnormality is recognized in the failure diagnosis of GPS receiver 13, the vehicle travels by autonomous navigation.
The estimated current position and azimuth gradually deviate due to the influence of the accumulated error between the gyro 21 and the speed sensor 22, and travel is controlled based on the deviated position and azimuth information. Therefore, as shown in FIG. Gradually deviates from the target traveling route 4 by the deviation. At this time, the current position is estimated on the basis of the difference between the measured value of the roadside belt distance L by the roadside belt distance measuring device 15 and the reference distance L0, and the estimated current position based on the roadside belt distance L and the autonomous navigation are used. The position is corrected by sensor fusion with the estimated current position to obtain a more reliable current position, and the azimuth is corrected based on the correction amount of the current position. When the roadside zone 5 is a bank or a wall on an ordinary substantially straight or gentle curve traveling road, based on the current position obtained above, only the lateral direction with respect to the traveling direction is used. Perform position correction. The reason is that in such a case, the position error in the traveling direction is smaller than that in the lateral direction, and the lateral error is more important for automatic traveling, but the traveling direction error is not a problem. Because there are many. If the measured roadside belt distance L is too small or too large, it is determined that the vehicle has deviated extremely from the travel route 4 and an emergency stop is performed before the correction is performed.

(When there is no bank, or when the traveling direction error is a problem) a) When the area R without the bank (see FIG. 7) continues over a long distance b) When turning at an intersection or turning sharply Case c) In the vicinity of a stop position such as a hopper, a dumping site, or a loading site,
When Positional Accuracy in Traveling Direction is Required In these cases, as shown in FIG. 7, a reflector 25 such as a reflective pole or signboard for positional correction in the traveling direction in which the installation coordinates are accurately measured in advance is used. Installed as roadside zone 5. The distance to the reflector 25 whose coordinates are known is determined by the roadside distance measuring device 1.
5, the current position is accurately estimated based on the measured distance, and the traveling direction position can be accurately corrected together with the lateral position. In order to determine the current position from the value measured by the roadside belt distance measuring device 15, the reflection plate 25
, The traveling direction of the vehicle at the time of measurement, and the transmission direction of the laser or ultrasonic wave from the roadside distance measuring device 15 with respect to this traveling direction. Therefore, if positioning by GPS becomes impossible or abnormal in such a place, or if abnormality is recognized in autonomous navigation, positioning data of the current position by the roadside distance measuring device 15 and positioning by GPS are used. The sensor fusion with the positioning data of the normal position, whichever is the position determined by the autonomous navigation computing unit 14, and the sensor fusion, the traveling direction position accuracy can be increased, and a highly reliable position can be measured. As a result, it is possible to reliably steer along the traveling route 4, to accurately turn at an intersection or a sharp turning point, and to improve stopping accuracy at a stopping position.

(Sensor fusion of roadside zone measurement and GPS positioning when autonomous navigation is abnormal) When an abnormality of autonomous navigation arithmetic unit 14 due to failure of gyro 21 or speed sensor 22 or the like is recognized, depending on the abnormal form, As described above, the measuring means are switched to the measuring means forming a redundant system. a) When all the mounted speed sensors are abnormal The GPS speed is used instead of the speed sensor 22. G
The PS speed data is G every predetermined time (for example, 0.5 seconds).
Output from the PS receiver 13. b) When there is an abnormality related to the gyro The GPS attitude angle is used instead of the gyro 21. G
The PS attitude angle is an estimated value of the attitude angle output from the GPS receiver 13 or the attitude angle (in this case, the moving direction) based on the position and movement trajectory of the GPS antenna 12a.

Since the above-mentioned GPS speed is calculated based on the moving distance at every predetermined time, the accuracy at a low speed may be low. In addition, since the data output time interval does not become smaller than the GPS sampling time, the autonomous navigation is used. There is a case where the accuracy of the speed measurement is deteriorated. In addition, since the above GPS attitude angle is calculated from a change in the GPS antenna position,
Although the accuracy is high to some extent when the vehicle goes straight, there is a disadvantage that the accuracy is deteriorated when the vehicle turns, for example, because it does not match the vehicle attitude angle (that is, the traveling direction). Therefore, the position measurement accuracy based on the GPS speed and the GPS attitude angle is lower than that of the autonomous navigation. Therefore, when the vehicle is guided by pseudo autonomous navigation using the GPS data, the trajectory 7a is shifted as shown in FIG. Go. At this time, the current position calculated based on the actually measured distance L to the roadside zone 5 (including the bank, the above-described reflector 25, and the like) by the roadside zone distance measuring device 15 and the GPS speed and / or GPS attitude angle The estimation accuracy of the pseudo autonomous navigation is estimated from the amount of deviation from the current position measured by the above, and the current position is corrected based on the GPS speed and / or GPS attitude angle based on the estimated measurement accuracy. Then, a more reliable current position is obtained by sensor fusion of the corrected current position and the current position obtained by GPS positioning, and the azimuth is corrected based on the correction amount of the current position. If the difference between the measured value of the roadside belt distance and the reference distance L0 becomes extremely large before the above correction is performed, the vehicle deviates from the target traveling route 4 and the unmanned vehicle 10 is urgently stopped. Let it do.

(About bank stability determination)
When performing sensor fusion and abnormality detection based on roadside belt distance measurement, it is important to first consider the uniformity (gently changing) and continuity (interruption) when the measurement target is a bank or wall. This is true on the premise that when the evaluation result is always good (when it is determined to be stable), the positioning data based on the roadside distance measurement is used. That is, in areas where the roadside zone 5 is unstable, such as discontinuous breaks in areas where a bank or wall is previously installed, or a change greater than a preset change rate, It is determined that an abnormality has occurred, and the roadside zone distance measured in an area having such an abnormality is regarded as abnormal data. Second, in the case of a reflection pole, the reflection intensity of the reflection pole is equal to or higher than a predetermined value, and a signboard (a signboard having a predetermined shape).
In the case of (1), it is important that the characteristics of the reflection plate 25 can be detected by characterizing the shape, the installation position, the installation distance, and the like, so that the reflection plate 25 can be clearly distinguished from other objects. The determination of the stability of the bank or the like is performed, for example, by determining the relative amount to the vehicle of the bank calculated from the amount of change (difference value) of the roadside belt distance from the previous measurement value, the rate of change, or the measurement distance at a plurality of positions (that is, Evaluate based on parameters such as bank inclination). At this time, when the obtained parameter is greater than or equal to a predetermined allowable value (for example, an allowable value is set in advance for the roadside zone distance change rate obtained based on the terrain information GI), a change in the bank or the like is made. It is determined that it is abrupt and unstable, and the positioning data based on the roadside distance measurement is not used. Further, the stability determination evaluation result is used as a determination material for calculating an estimated accuracy value of the roadside belt distance measurement, and may be reflected in the weighting at the time of sensor fusion. Note that the method is not limited to the above method, and a stable area and an unstable area or an area without a bank are measured in advance and stored as terrain information GI on the vehicle control system side in the fixed base station 1, and the unmanned vehicle 10 May input the terrain information GI by communication from the vehicle control system.

The processing procedure of the sensor fusion of the position measuring unit 11 described above will be described with reference to the flowcharts shown in FIGS. First, in step S1, the GPS
Information related to GPS navigation is input from the receiver 13 to perform a failure diagnosis. Next, in step S2, the autonomous navigation computing unit 1
Input information of autonomous navigation from 4 and perform fault diagnosis. Next, in step S3, information related to roadside belt distance measurement is input from the roadside belt distance measurement device 15, and failure diagnosis is performed. Further, in step S4, the travel route data CD and the terrain information GI are input from the vehicle control system of the fixed base station 1 by communication, and a failure diagnosis is performed.

Then, in step S5, the traveling route data C
It is checked whether D is normal, that is, whether the communication of the traveling route data CD is normal and whether the received data is continuous. If it is normal, it is determined in step S7 whether the terrain information GI is normal.
That is, it is checked whether the communication of the terrain information GI is normal and whether the received data is normal (for example, whether the change amount of the coordinate value of the bank, the change rate is within a predetermined value, etc.). If the terrain information GI is normal, it is checked in step S8 whether or not the fault is related to the roadside zone distance measurement, and if it is normal, the stability of the roadside zone 5 obtained based on the roadside zone distance measurement data in step S9 is determined. Determine if it is good. If it is not normal in steps S7 and S8, and if the stability is not good in step S9, the roadside zone measurement impossible flag is set in step S10, and the process proceeds to step S11. If the traveling route data CD is not normal in step S5, a vehicle stop command is output to the traveling control unit 17 in step S6 to stop the vehicle.

If the stability is good in step S9, it is checked in step S11 whether the failure diagnosis result of the autonomous navigation is normal. If it is normal, next, in step S12, it is checked whether the failure diagnosis result of the GPS navigation is normal. If it is normal, it is further checked in step S13 whether the roadside zone measurement impossible flag is set. Thereafter, when the roadside zone measurement impossible flag is not set, the current position is calculated by the autonomous navigation, the GPS navigation and the roadside zone distance measurement in step S14, and a difference value between the current positions is obtained. In S15, it is checked whether or not each of the difference values is within a predetermined value. If the difference value is within the predetermined value, it is determined in step S16 that there is no divergence (mutual inconsistency) between the current position data, and GPS navigation, autonomous navigation, and roadside zone distance measurement are performed. The sensor fusion is performed based on the respective current position and direction data obtained by the above, and the corrected current position and direction are calculated. Then, in step 18, the corrected current position and direction are determined as the final vehicle position and direction, and the data is output to the traffic control unit 16, the traveling control unit 17, and the safety management unit 19. Traffic control unit 16, running control unit 17, and safety management unit 1
Reference numeral 9 performs predetermined control, traveling control, and safe traveling management based on the corrected current position and orientation, the target position, and the target traveling posture, respectively. Thereafter, the process returns to step S1 to repeat the processing.

If the difference values are not within the predetermined values in step S15, it is determined in step S17 that there is a deviation between the current position data, and a vehicle stop command is output to the traveling control unit 17 to stop the vehicle. Let it.

If the roadside zone measurement impossible flag is set in step S13, the current position and the azimuth are calculated by the autonomous navigation and the GPS navigation in step S19, and the difference between the two current positions is obtained. Step S2
It is checked whether this difference value is within a predetermined value at 0. If it is within the predetermined value, it is determined in step S21 that there is no divergence between the two current position data, and the respective current position and direction data obtained by the GPS navigation and the autonomous navigation are determined. The sensor fusion is performed on the basis of the above, the corrected current position and the azimuth are calculated, and thereafter, the process proceeds to the step S18. Step S2
If the difference is not equal to or smaller than the predetermined value, the process proceeds to step S17 to stop the vehicle.

If the failure diagnosis result of the GPS navigation is not normal in step S12, it is checked in step S22 whether the roadside zone measurement impossible flag is set. If the roadside zone measurement impossible flag is not set, step S22 is executed.
At 23, the current position is calculated by the autonomous navigation and the roadside distance measurement, and the difference between the two current positions is obtained. Next, in step S24, it is checked whether the difference value is within a predetermined value. If the difference value is within the predetermined value, it is determined in step S25 that there is no divergence between the two current position data, and the respective values obtained by the autonomous navigation and the roadside zone distance measurement are determined. The sensor fusion is performed based on the current position and direction data to calculate the corrected current position and direction. Thereafter, the process proceeds to step S18. If it is determined in step S24 that the difference value is not within the predetermined value, the process proceeds to step S17 to stop the vehicle.

If the roadside zone measurement impossible flag is set in step S22, the current position and azimuth are estimated and calculated by autonomous navigation in step S26.
In step 7, a command is output to the travel control unit 17 to decelerate and stop the vehicle based on the estimated current position and direction.

If the failure diagnosis result of the autonomous navigation is not normal in step S11, it is checked in step S28 whether the failure diagnosis result of the GPS navigation is normal. If it is normal, in step S29, the redundancy between the traveling speed and the traveling direction of the autonomous navigation is switched, that is, if it is diagnosed that all the input data relating to the traveling speed (when there are a plurality of sensors) or all the speed sensors are abnormal, If it is determined that the input data relating to the traveling direction is abnormal or the azimuth sensor is abnormal using the GPS speed as the data, autonomous navigation is performed using the GPS attitude as the traveling direction data.

In step S30, it is checked whether the roadside zone measurement impossible flag is set. If the roadside zone measurement impossible flag is not set, step S31 is executed.
The current position is calculated by autonomous navigation based on the measured data of the speed sensor and the direction sensor that has no failure and the substitute GPS speed and / or GPS attitude of the failed sensor, and the GPS navigation and the roadside zone are used. The current position is calculated by the distance measurement, and a difference value between the current positions is obtained. Next, in step S32, it is checked whether or not each of the difference values is within a predetermined value.
Navigation, autonomous navigation (the alternative GPS speed and / or GP
Based on the respective current position data obtained based on the S side posture) and the roadside belt distance measurement, sensor fusion is performed to calculate a corrected current position and orientation, and the process proceeds to step S18. If the respective difference values are not within the predetermined values in step S32, the process proceeds to step S17.

If the roadside zone measurement disable flag is set in step S30, a vehicle deceleration stop command is output in step 34 to start deceleration of the vehicle. , The substitute GPS
The alternative autonomous navigation based on the speed and / or the GPS attitude is continued, and the process proceeds to step 18 to perform a predetermined process.

If the failure diagnosis result of the GPS navigation is not normal in step S28, it is checked in step S36 whether the roadside zone measurement impossible flag is set. If the roadside zone measurement impossible flag is not set, step S36 is executed.
At 37, a vehicle deceleration stop command is output to decelerate and stop the vehicle. In addition, for steering guidance until the vehicle stops, a command is output to control traveling based on the measured value of the roadside zone measurement and the terrain information GI data. Next, in step S38,
Until the vehicle stops, the data of the autonomous navigation data of the driving distance and the driving direction in which no abnormality has occurred and the data of the one in which the abnormality has been recognized, the past normal vehicle speed and / or vehicle attitude, The autonomous navigation based on the vehicle speed and / or the traveling direction data estimated from the vehicle trajectory, the current steering angle, and the like, and the position correction calculation based on the roadside distance measurement are performed. If both the traveling distance and traveling direction of the autonomous navigation data are abnormal, the vehicle speed and traveling direction data estimated from the past normal vehicle speed and vehicle attitude, vehicle trajectory, current steering angle, etc. And a position correction calculation based on roadside belt distance measurement based on autonomous navigation based on prediction until vehicle stop based on the vehicle.

If the roadside zone measurement impossible flag is set in step S36, the vehicle is emergency stopped by applying sudden braking in step S39. Next, in step 40,
Until the vehicle stops, the autonomous navigation data of the traveling distance and traveling direction in which the abnormality has not occurred and the data in which the abnormality has been recognized are compared with the past normal vehicle speed and / or vehicle attitude, vehicle Autonomous navigation is performed by estimating from the trajectory, the current steering angle, and the like. In addition, when both the traveling distance and the traveling direction in the autonomous navigation data are abnormal, prediction of the vehicle speed and vehicle attitude, vehicle trajectory, vehicle trajectory, current steering angle, etc., in the past and normal, in both cases, until the vehicle stops. Implement autonomous navigation based on it.

FIG. 11 is a travel locus diagram showing the results of an actual machine test performed using the travel control device according to the present invention. In FIG. 5, reference numeral 5a denotes a roadside zone 5 of an area in which the result of the stability determination is good when the vehicle travels along the travel route 4 and the distance to the bank is measured, and 5b (shown by a thick line) indicates the stability determination. The roadside zone 5 of the area where the result was NG is shown. When the vehicle travels along the traveling route 4 provided with the roadside zone 5 only by the autonomous navigation (assuming that positioning by the GPS navigation is abnormal or impossible), the vehicle travels as a trajectory 4a. It has been shown. Then, at this time, when the roadside zone stability determination result is good and the current position is calculated by the sensor fusion between the roadside zone distance measurement and the autonomous navigation, the vehicle travels without deviating from the travel route 4 as shown in the figure. I confirmed that I would run. From this, it is understood that the positioning accuracy of the vehicle position can be improved by the sensor fusion with the roadside distance measurement.

As described above, the present invention has the following effects. (1) Based on the measured value of the distance to the roadside zone 5 constituted by the road shoulder, the signboard, the reflector 25, etc., the positioning position (and the azimuth if necessary) by GPS navigation and / or the positioning position by autonomous navigation Since the current position and direction are calculated by correcting at least one of the directions with sensor fusion, a decrease in position measurement accuracy due to GPS positioning errors or autonomous navigation positioning errors is suppressed, and a more reliable vehicle position is obtained. Positioning is possible. (2) In addition, the deterioration of the communication environment such as an increase in the delay time of the correction data from the fixed base station 1 and the deterioration of the reception environment such as the visibility of the sky, the satellite constellation, and the multipath cause the GPS navigation to be unable to perform positioning or positioning. If the accuracy is deteriorated, positioning by autonomous navigation is disabled, or any of the positioning data in both navigation is abnormal, the positioning data and route by the normal GPS navigation or autonomous navigation are used. A reliable current position and azimuth can be calculated by sensor fusion based on the measured value of the side zone distance. Therefore, even if the above-described abnormality occurs, the guided traveling can be reliably controlled, and the operation rate of the unmanned vehicle can be increased. As a result, a highly reliable traveling system for unmanned vehicles can be configured.

(3) The roadside belt distance measured in the area where the roadside belt 5 is unstable is regarded as abnormal data, and when the stability of the roadside belt 5 is recognized, the positioning data obtained by the roadside belt distance measurement is used. . Based on the measured value of the roadside zone distance, sensor fusion is used to correct at least one of the positioning position by GPS navigation (and the azimuth if necessary) and / or the positioning position and azimuth by autonomous navigation. When the position 5 is unstable, it is possible to prevent a decrease in positioning accuracy caused by performing the correction. Further, when the stability of the roadside zone 5 is recognized, it is checked whether or not there is a deviation (data abnormality) between the positioning data based on the measured value of the roadside zone distance, so that it is possible to make a correct determination. As a result, the automatic traveling can be continued without being affected by the unstable state of the roadside zone 5, and the operating rate of the unmanned vehicle can be increased.

(4) Since the corrected position is calculated by the sensor fusion according to the positioning accuracy of each of the GPS receiver 13, the autonomous navigation calculator 14, and the roadside distance measuring device 15, the position and the position of the more reliable The direction can be determined. Therefore, the guided traveling of the unmanned vehicle 10 can be reliably controlled.

(5) Since the vehicle position and direction are measured based on the terrain information GI and the roadside zone distance, for example, both the GPS navigation and the autonomous navigation are temporarily disabled, Even if an abnormality is found, the vehicle can automatically decelerate without stopping immediately and perform guided driving only by measuring the roadside belt distance. Therefore, the frequency of the guidance stop of the unmanned vehicle is reduced and the operation rate is increased, so that a highly reliable traveling system can be configured. (6) The terrain information GI stores a fixed value of the roadside zone distance (reference distance L0), the coordinate position of the roadside zone 5 corresponding to the traveling route 4, or the roadside zone distance, depending on the state of the roadside zone 5. Accordingly, the traveling control device can cope with the traveling route 4 under various terrain conditions, and a general-purpose traveling system can be configured. (7) In an area where the traveling direction of the vehicle is important, for example, in an area without a bank or near a sharp turning position or a stop position, by installing the reflector 25 as a roadside zone, highly accurate positioning can be performed. And reliability can be further improved.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of an unmanned vehicle control device of the present invention.

FIG. 2 is an explanatory diagram of a traveling road having a uniform road width.

FIG. 3 is an explanatory diagram of a traveling path when the distance between the traveling path and the roadside zone is uneven.

FIG. 4 is an explanatory diagram of sensor fusion between GPS and autonomous navigation.

FIG. 5 is an explanatory diagram of a positioning data abnormality detection method based on a roadside distance measurement value.

FIG. 6 is an explanatory diagram of sensor fusion of roadside zone measurement and autonomous navigation.

FIG. 7 is an explanatory view of a traveling path when a reflector is installed.

FIG. 8 is an explanatory diagram of sensor fusion of roadside zone measurement and positioning by GPS.

FIG. 9 is an example of a flowchart (first half) illustrating a processing procedure of sensor fusion.

FIG. 10 is an example of a flowchart (second half) representing a processing procedure of sensor fusion.

FIG. 11 is a traveling locus diagram showing actual machine test results.

FIG. 12 is a configuration diagram of a traveling system of an unmanned vehicle.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Fixed base station, 4 ... Running route, 5 ... Roadside zone, 7, 7a
... track, 10 ... unmanned vehicle, 11 ... position measurement unit, 13 ... G
PS receiver, 14: Autonomous navigation calculator, 15: Roadside zone distance measuring device, 16: Traffic control unit, 17: Travel control unit, 18 ...
Actuator, 19: Safety management unit, 21: Gyro,
22: speed sensor, 23: laser distance meter, 25: reflector.

Claims (3)

[Claims] 1. A GPS receiver (13) that receives a GPS signal to measure an absolute coordinate position of a vehicle, and measures a traveling direction and a traveling distance of the vehicle. Autonomous navigation calculator (14) for measuring relative coordinate position and direction, GPS receiver (13) and autonomous navigation calculator (14)
A position measuring unit (11) for calculating the current position and azimuth of the vehicle based on the respective positioning results, and a preset traveling route
A traveling control unit (1) that compares the position and orientation of (4) with the current position and orientation calculated by the position measuring unit (11) and controls the vehicle traveling so as to reduce the respective deviation values.
7) a travel control device for an unmanned vehicle, comprising: a roadside zone distance measuring device (15) for measuring a distance from the vehicle to a roadside zone (5) provided beside the travel route (4); The measuring unit (11) is based on the roadside distance measured by the roadside distance measuring device (15), based on the positioning position of the GPS receiver (13), and / or the positioning position of the autonomous navigation calculator (14). A travel control device for an unmanned vehicle, wherein at least one of the azimuths is corrected to obtain a current position and an azimuth. 2. The travel control device for an unmanned vehicle according to claim 1, wherein the position measuring unit (11) is configured to determine a roadside zone from a measured value of a roadside zone distance.
(5) A travel control device for an unmanned vehicle, wherein the stability is determined, and when the stability is good, the correction is performed based on the roadside belt distance. 3. The travel control device for an unmanned vehicle according to claim 1, wherein the correction of the position measuring unit based on a roadside belt distance is performed by a GP.
To be performed by sensor fusion according to the estimated positioning accuracy of the S receiver (13) and / or the estimated positioning accuracy of the autonomous navigation calculator (14) and the estimated positioning accuracy of the roadside distance measuring device (15). A travel control device for an unmanned vehicle.